[Met_help] [rt.rap.ucar.edu #46779] History for Mode Analysis Properties
RAL HelpDesk {for John Halley Gotway}
met_help at ucar.edu
Tue May 24 11:30:06 MDT 2011
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
I have been using the MET software for the past few months and building case studies that use the mode analysis tool from MET. Currently, I am trying to build a study that tracks wind speed as it changes over the forecast hours. I was wondering how you would go about creating objects that look at the wind speed and using the mode analysis to plot the forecast and observation objects. Do I type in something that denotes the u and v components of the wind or do I have to create the total calculated wind speed in the GRIB file prior to running the mode analysis tool? Can tracking wind speed even be done with mode analysis?
Thank you for your help.
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #46779] Mode Analysis Properties
From: John Halley Gotway
Time: Tue May 10 12:13:02 2011
Kris,
I went ahead and created a met_help ticket to track your question.
I'd suggest starting by thinking about how you'd like to use the MODE
tool to define objects. If you pass a GRIB file containing U and V to
the MODE tool and set the "fcst_field = "WIND";" in the
MODE config file, MODE should derive the wind speed field for you. Go
ahead and try that to see if the wind speed objects you define look
reasonable to you.
Randy Bullock and Tressa Fowler (in the DTC at NCAR) did some work
last year applying MODE to other wind derived fields - like vorticity
and divergence. They found that those fields worked pretty
well with an object-based approach. And they found that wind speed
worked reasonably well too.
Once you run MODE on these wind speed objects over many cases through
time, you could use the MODE-Analysis tool to get some sense of how
these objects change in time.
Hopefully that will help get you started.
Thanks,
John Halley Gotway
met_help at ucar.edu
On 05/10/2011 12:05 PM, RAL HelpDesk {for John Halley Gotway} wrote:
>
> Tue May 10 12:05:28 2011: Request 46779 was acted upon.
> Transaction: Ticket created by johnhg
> Queue: met_help
> Subject: Mode Analysis Properties
> Owner: Nobody
> Requestors: kristopher.pelletier at offutt.af.mil
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46779 >
>
>
> I have been using the MET software for the past few months and
building case studies that use the mode analysis tool from MET.
Currently, I am trying to build a study that tracks wind speed as it
changes over the forecast hours. I was wondering how you would go
about creating objects that look at the wind speed and using the mode
analysis to plot the forecast and observation objects. Do I type in
something that denotes the u and v components of the wind or do I have
to create the total calculated wind speed in the GRIB file prior to
running the mode analysis tool? Can tracking wind speed even be done
with mode analysis?
>
> Thank you for your help.
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #46779] Mode Analysis Properties
From: John Halley Gotway
Time: Tue May 10 15:05:03 2011
Kris,
Well if you have observations (or some model analysis) to which you
want to compare the wind forecast, I'd set: obs_field = "WIND";
If you're really not interested in comparing the forecast objects to
any observations, I would just pass the forecast file in on the
command line in both the forecast spot and observation spot. So
just compare the field to itself. And if you want MODE to run faster,
you could disable matching and merging by setting fcst_merge_flag,
obs_merge_flag, and match_flag all to 0.
Then you'd only be interested in the "simple single forecast objects"
in MODE-Analysis.
Clear as mud?
John
On 05/10/2011 12:28 PM, Pelletier, Kristopher Civ USAF AFWA 16 WS/WXN
wrote:
> John,
>
> I'm putting WIND in the forecast field, but I was wondering what I
put in the obs field. Do I leave it blank or put in the components or
what? Thanks for all the help.
>
>
> //SIGNED//
> KRIS PELLETIER, Meteorology SCEP
> Air Force Weather Agency 2WXG/16WS/WXN
> Offutt AFB, NE
> 402.232.7207
>
>
> -----Original Message-----
> From: John Halley Gotway [mailto:johnhg at ucar.edu]
> Sent: Tuesday, May 10, 2011 1:13 PM
> To: met_help at ucar.edu; Pelletier, Kristopher Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #46779] Mode Analysis Properties
>
> Kris,
>
> I went ahead and created a met_help ticket to track your question.
>
> I'd suggest starting by thinking about how you'd like to use the
MODE tool to define objects. If you pass a GRIB file containing U and
V to the MODE tool and set the "fcst_field = "WIND";" in the
> MODE config file, MODE should derive the wind speed field for you.
Go ahead and try that to see if the wind speed objects you define look
reasonable to you.
>
> Randy Bullock and Tressa Fowler (in the DTC at NCAR) did some work
last year applying MODE to other wind derived fields - like vorticity
and divergence. They found that those fields worked pretty
> well with an object-based approach. And they found that wind speed
worked reasonably well too.
>
> Once you run MODE on these wind speed objects over many cases
through time, you could use the MODE-Analysis tool to get some sense
of how these objects change in time.
>
> Hopefully that will help get you started.
>
> Thanks,
> John Halley Gotway
> met_help at ucar.edu
>
> On 05/10/2011 12:05 PM, RAL HelpDesk {for John Halley Gotway} wrote:
>>
>> Tue May 10 12:05:28 2011: Request 46779 was acted upon.
>> Transaction: Ticket created by johnhg
>> Queue: met_help
>> Subject: Mode Analysis Properties
>> Owner: Nobody
>> Requestors: kristopher.pelletier at offutt.af.mil
>> Status: new
>> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46779 >
>>
>>
>> I have been using the MET software for the past few months and
building case studies that use the mode analysis tool from MET.
Currently, I am trying to build a study that tracks wind speed as it
changes over the forecast hours. I was wondering how you would go
about creating objects that look at the wind speed and using the mode
analysis to plot the forecast and observation objects. Do I type in
something that denotes the u and v components of the wind or do I have
to create the total calculated wind speed in the GRIB file prior to
running the mode analysis tool? Can tracking wind speed even be done
with mode analysis?
>>
>> Thank you for your help.
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #46779] Mode Analysis Properties
From: John Halley Gotway
Time: Mon May 16 11:09:18 2011
Kris,
I apologize for the delay in getting back with you. The end of last
week was very busy.
So you really want to run MODE to simply define objects in a single
field. You aren't really interested in comparing the forecast objects
to the observed objects. You will need to specify something
in the "obs_field" - otherwise MODE will complain. One approach you
could try would be to compare the same field to itself - and setup the
configuration to define forecast objects the way you want
them to be, but define zero observation objects.
For your task, you may find the following configuration settings
helpful:
fcst_field = "WIND/Z10"; // To look at 10-meter winds for
example
obs_field = "WIND/Z10"; // Here we're comparing the same
field to itself
fcst_conv_radius = 5; // Or whatever amount of smoothing
you'd like
obs_conv_radius = 5; // Set to 0 since we don't want to
create any observation objects anyway
fcst_conv_thresh = "ge10.0"; // Or whatever threshold you'd
like
obs_conv_thresh = "ge100000.0"; // Setting this very high will
ensure that no observation objects are defined
fcst_merge_flag = 0; // Set these 3 flags to 0 to avoid
wasting time merging/matching objects when we know we won't have any
observed objects anyway
obs_merge_flag = 0;
match_flag = 0;
I hope that helps. Just let us know if any more questions arise.
Thanks,
John Halley Gotway
met_help at ucar.edu
On 05/10/2011 03:05 PM, John Halley Gotway wrote:
> Kris,
>
> Well if you have observations (or some model analysis) to which you
want to compare the wind forecast, I'd set: obs_field = "WIND";
>
> If you're really not interested in comparing the forecast objects to
any observations, I would just pass the forecast file in on the
command line in both the forecast spot and observation spot. So
> just compare the field to itself. And if you want MODE to run
faster, you could disable matching and merging by setting
fcst_merge_flag, obs_merge_flag, and match_flag all to 0.
>
> Then you'd only be interested in the "simple single forecast
objects" in MODE-Analysis.
>
> Clear as mud?
>
> John
>
> On 05/10/2011 12:28 PM, Pelletier, Kristopher Civ USAF AFWA 16
WS/WXN wrote:
>> John,
>>
>> I'm putting WIND in the forecast field, but I was wondering what I
put in the obs field. Do I leave it blank or put in the components or
what? Thanks for all the help.
>>
>>
>> //SIGNED//
>> KRIS PELLETIER, Meteorology SCEP
>> Air Force Weather Agency 2WXG/16WS/WXN
>> Offutt AFB, NE
>> 402.232.7207
>>
>>
>> -----Original Message-----
>> From: John Halley Gotway [mailto:johnhg at ucar.edu]
>> Sent: Tuesday, May 10, 2011 1:13 PM
>> To: met_help at ucar.edu; Pelletier, Kristopher Civ USAF AFWA 16
WS/WXN
>> Subject: Re: [rt.rap.ucar.edu #46779] Mode Analysis Properties
>>
>> Kris,
>>
>> I went ahead and created a met_help ticket to track your question.
>>
>> I'd suggest starting by thinking about how you'd like to use the
MODE tool to define objects. If you pass a GRIB file containing U and
V to the MODE tool and set the "fcst_field = "WIND";" in the
>> MODE config file, MODE should derive the wind speed field for you.
Go ahead and try that to see if the wind speed objects you define look
reasonable to you.
>>
>> Randy Bullock and Tressa Fowler (in the DTC at NCAR) did some work
last year applying MODE to other wind derived fields - like vorticity
and divergence. They found that those fields worked pretty
>> well with an object-based approach. And they found that wind speed
worked reasonably well too.
>>
>> Once you run MODE on these wind speed objects over many cases
through time, you could use the MODE-Analysis tool to get some sense
of how these objects change in time.
>>
>> Hopefully that will help get you started.
>>
>> Thanks,
>> John Halley Gotway
>> met_help at ucar.edu
>>
>> On 05/10/2011 12:05 PM, RAL HelpDesk {for John Halley Gotway}
wrote:
>>>
>>> Tue May 10 12:05:28 2011: Request 46779 was acted upon.
>>> Transaction: Ticket created by johnhg
>>> Queue: met_help
>>> Subject: Mode Analysis Properties
>>> Owner: Nobody
>>> Requestors: kristopher.pelletier at offutt.af.mil
>>> Status: new
>>> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46779 >
>>>
>>>
>>> I have been using the MET software for the past few months and
building case studies that use the mode analysis tool from MET.
Currently, I am trying to build a study that tracks wind speed as it
changes over the forecast hours. I was wondering how you would go
about creating objects that look at the wind speed and using the mode
analysis to plot the forecast and observation objects. Do I type in
something that denotes the u and v components of the wind or do I have
to create the total calculated wind speed in the GRIB file prior to
running the mode analysis tool? Can tracking wind speed even be done
with mode analysis?
>>>
>>> Thank you for your help.
------------------------------------------------
More information about the Met_help
mailing list